À la fin de ce chapitre, vous
Variables relatives à un tout.
Sachant qu’une journée a 24 heures et que je passe 8 heures au travail, il reste implicitement 16 h hors du travail.
Si je segmente la journée en plusieurs tâches au travail et plusieurs tâches hors du travail, les statistiques au travail ne devraient être influencées par le fait que les heures soient exprimées en proportion de mon temps de travail (/8h) ou en proportion de mon temps total (/24h).
Les statistiques ne devraient pas admettre des proportions négatives ou dépassant 1 (ou 100%, ou 24h).
\[alr_j = log \left( \frac{x_j}{x_{ref}} \right)\] \[ clr_i = log \left( \frac{x_i}{g \left( x \right)} \right) \] \[ ilr_j = \sqrt{\frac{n_j^+ n_j^-}{n_j^+ + n_j^-}} log \left( \frac{g \left( c_j^+ \right)}{g \left( c_j^+ \right)} \right) \]
sbp <- matrix(c(1, 1,-1,
1,-1, 0),
byrow = TRUE,
ncol = 3)
CoDaDendrogram(comp, V = gsi.buildilrBase(t(sbp)))[Argile | Limon,Sable], [Limon | Sable]
Rechercher une station par coordonnées
library("weathercan")
station_site <- stations_search(coords = c(45.35, -71.90), dist = 20, interval = "hour")
station_site## # A tibble: 4 x 14
## prov station_name station_id climate_id WMO_id TC_id lat lon elev
## <fct> <chr> <fct> <fct> <fct> <fct> <dbl> <dbl> <dbl>
## 1 QC LENNOXVILLE 5397 7024280 71611 WQH 45.4 -71.8 181
## 2 QC SHERBROOKE 48371 7028123 71610 YSC 45.4 -71.7 241.
## 3 QC SHERBROOKE A 5530 7028124 71610 YSC 45.4 -71.7 241.
## 4 QC SHERBROOKE A 30171 7028126 <NA> GSC 45.4 -71.7 241.
## # … with 5 more variables: tz <chr>, interval <chr>, start <int>,
## # end <int>, distance <dbl>
mont_bellevue <- weather_dl(station_ids = c(5397, 48371), start = "2019-01-01", end = "2019-01-07")
mont_bellevue %>%
ggplot(aes(x = time, y = temp)) +
geom_line(aes(colour = station_name))library("soiltexture")
rand_text <- TT.dataset(n=100, seed.val=29)
png("images/soiltexture_1.png", width = 600, height = 600, res = 90)
TT.plot(class.sys = "USDA.TT",
tri.data = rand_text,
col = "blue")
dev.off()## png
## 2
## [1] "ALi" "ALi" "L" "L" "ALo" "LS" "ALo" "A" "LLi" "LSA"
## Loading 'meta' package (version 4.9-4).
## Type 'help(meta)' for a brief overview.
## Parsed with column specification:
## cols(
## author = col_character(),
## Ne = col_double(),
## Me = col_double(),
## Se = col_double(),
## Nc = col_double(),
## Mc = col_double(),
## Sc = col_double()
## )
## SMD 95%-CI %W(fixed) %W(random)
## 1 -0.5990 [-1.3300; 0.1320] 3.5 5.7
## 2 -0.9518 [-1.6770; -0.2266] 3.6 5.7
## 3 -0.5909 [-1.6301; 0.4483] 1.7 4.1
## 4 -0.7064 [-1.7986; 0.3858] 1.6 3.9
## 5 -0.2815 [-0.6076; 0.0445] 17.6 8.1
## 6 -0.5375 [-1.0816; 0.0065] 6.3 6.8
## 7 -1.3204 [-2.1896; -0.4513] 2.5 4.9
## 8 -0.4800 [-1.3514; 0.3914] 2.5 4.9
## 9 0.0918 [-0.2549; 0.4385] 15.6 8.0
## 10 -3.2433 [-4.2035; -2.2831] 2.0 4.5
## 11 0.0000 [-0.7427; 0.7427] 3.4 5.6
## 12 -0.7061 [-1.2020; -0.2102] 7.6 7.1
## 13 -0.4724 [-1.2537; 0.3089] 3.1 5.4
## 14 -0.1849 [-0.5071; 0.1373] 18.0 8.2
## 15 -0.0265 [-0.6045; 0.5515] 5.6 6.6
## 16 -1.1648 [-2.0828; -0.2468] 2.2 4.7
## 17 -0.2127 [-0.9651; 0.5397] 3.3 5.6
##
## Number of studies combined: k = 17
##
## SMD 95%-CI z p-value
## Fixed effect model -0.3915 [-0.5283; -0.2548] -5.61 < 0.0001
## Random effects model -0.5858 [-0.8703; -0.3013] -4.04 < 0.0001
##
## Quantifying heterogeneity:
## tau^2 = 0.2309; H = 1.91 [1.50; 2.43]; I^2 = 72.5% [55.4%; 83.1%]
##
## Test of heterogeneity:
## Q d.f. p-value
## 58.27 16 < 0.0001
##
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Hedges' g (bias corrected standardised mean difference)
L’entreprise RStudio a créé shiny pour permettre le développement d’applications R.
À la fin de ce chapitre, vous